a layer on top of other apps/specifications (bids-app, flywheel gears, boutiues) which provide formalization of the INTERFACES but do not really give an idea on what/how much of resources the app would need given the job parameters: dataset, parametrization of the app, etc.
Examples of such "app specific" knowledge specifications:
brainlife.io - then encode them as ready-to-use #PBS options within their run scripts as in https://github.com/brainlife/app-networkneuro/blob/master/main#L2 . @soichih, if you are reading this -- are all those clusters using PBS or do you have some mapping from those options into SLURM or other scheduler?
kubernetes podspec immediately came to @asmacdo mind (remove if that is not true)
a layer on top of other apps/specifications (bids-app, flywheel gears, boutiues) which provide formalization of the INTERFACES but do not really give an idea on what/how much of resources the app would need given the job parameters: dataset, parametrization of the app, etc.
Examples of such "app specific" knowledge specifications:
#PBSoptions within theirrunscripts as in https://github.com/brainlife/app-networkneuro/blob/master/main#L2 . @soichih, if you are reading this -- are all those clusters using PBS or do you have some mapping from those options into SLURM or other scheduler?kubernetes podspec immediately came to @asmacdo mind (remove if that is not true)